posit::glimpse() Newsletter – December 2024

New Products and Recent Releases

Positron

Positron is Posit’s new data science IDE.  It is an extensible, polyglot tool for writing code and exploring data in both R and Python. Check out how to get started using Positron or explore our user guides on topics such as the data explorer and choosing interpreters. Positron leverages the new ARK framework, an R kernel for Jupyter applications which is compatible with any frontend implementing the Jupyter protocol. Our FAQs cover some common questions you may have.  If that question is “What does this mean for RStudio?“, know that we are absolutely committed to maintaining and updating the RStudio IDE.

Packages to help leverage AI and Large Language Models in your work: Shiny Assistant, Elmer, Chatlas, PAL and Mall

• Shiny Assistant: Shiny Assistant brings AI-powered development tools to Shiny, helping developers accelerate their workflows with code generation capabilities. The Assistant integrates seamlessly with Shiny for R and Python, offering support for prototyping and building interactive web applications more efficiently. 
• elmer (R) and chatlas (Python): These packages make it easy to program with LLMs  from R and python. They provide streaming output, tool/function calling, and structured data extraction, allowing you to easily prototype new LLM-powered solutions. 
• pal: A library of ergonomic LLM assistants designed to help you complete repetitive, hard-to-automate tasks quickly.
• mall: Lightweight R and Python packages that make it easy to apply local (ollama) models to your data, to do things like extract, summarize, translate and analyze text. Read about it or visit the package website and give it a try.

brand.yml

brand.yml is a new feature recently released with Quarto v1.6 and Shiny for Python.  It is a simple, portable YAML file that codifies your company’s brand guidelines (colors, fonts, and logos) into a format that can be used to create consistent branded outputs — from web applications to printed reports to dashboards and presentations to (eventually) ggplot2 and plotnine plots. The goal is to make it easy and pleasant to maintain consistency across your organization’s applications and documents with minimal effort.

great_tables, gt, and reactable

The latest update to the Python great_tables package (v0.14.0) introduces LaTeX output for PDF generation, continuing its focus on creating beautifully styled tables in Python. Its R counterpart, gt (v0.11.1), released a few months ago also enhances LaTeX support while adding new datasets. Joining the family is the newly released Python port of reactable, designed for interactively viewing DataFrames.

Posit Connect Cloud

Connect Cloud simplifies the deployment and sharing of your data science outputs such as Shiny, Streamlit, and Dash as well as Quarto, R Markdown, and Jupyter.  Simply point to code in public GitHub repositories to deploy on Connect Cloud without worrying about infrastructure. For more details, create a free account and review workflow examples in the docs.

Our on-prem equivalent, Posit Connect 2024.11.0 also received a recent update: It adds push-button publishing for VS Code & Positron, plus APIs for auditing packages & updates to OAuth integrations.

Community and Learning 

Posit Conf 2024 Videos

The recorded sessions from Posit Conf 2024 are now available and include quite a few on technical publishing with Quarto and building data-driven applications with Shiny with a strong focus on leveraging AI/LLM.  The talk on Positron is especially insightful for those interested in upcoming features and the future direction of Posit’s newest data science IDE.

Open Source Contests –  These events promote hands-on learning, sharing with reproducible examples, and celebrating great work by the community.

• Shiny Contest – Over 100 Shiny developers shared their fun and practical Shiny apps, meet the winners here.
• Great Tables Contest – See community-contributed data display tables — beautiful & and highly stylized — built entirely from R or Python. Highlights here.
• Plotnine Contest – Hassan Kibirige organized the first Plotnine community sharing event showing off the amazing plots possible with the grammar of graphics for Python.

IN PROGRESS Closeread Prize – Scrollytelling with Quarto, Deadline Dec 15 – If you’ve ever wanted to create impactful scrollytelling stories like the ones you see in online news — all based on code and markdown in a Quarto doc — check out this community event.

Video Workshops

The Posit open-source team has been working on a series of free workshop-style videos.  Check out Creating Quarto Dashboards with Mine and Quarto Websites with Charlotte and Emil. Keep an eye out on our YouTube channel for a series on great_tables coming soon. 

Tidymodel prediction workflows with orbital & Snowflake

Last week we highlighted how to fit and deploy R models directly in Snowflake and drastically reduce model runtime. You can view the recording here.

Upcoming Data Science Hangouts

Join us at our weekly gathering place for the data science community, held every Thursday at 12 ET. Leaders from DuckDB Labs, Merck, and Penguin Random House will join us to chat about career lessons, tools, workflows, and much more. Listen in, join the conversation, or ask questions anonymously!

Building Community on Bluesky

Bluesky is taking off! Please join us there with this handy Posit employees starter pack. 

If you want to sign up for future, quarterly issues of glimpse or to make sure you receive imminent and exciting posit::conf(2025) news, we appreciate you subscribing here.

Until next time, we appreciate you including us on your  data science journey—here’s to learning, building, and growing together!